Computing infrastructure is one of new infrastructure, will be conductive to scientific and technological innovation, industrial upgrading and social governance. on the basis of the definition and classification of computing infrastructure, this paper expounds the important development trends of computing heterogeneity, computing and network coordination, ubiquitous computing power, green and low-carbon, analyzes challenges faced by computing infrastructure construction in China, and puts forward corresponding strategic advice for high-quality development of computing infrastructure from the aspects of industrial ecological construction, technical system research, overall resource layout, and green and low-carbon optimization.
Heterogeneous AI computing power operation platform is an important carrier for efficient and reliable processing and application of diversified data, which can meet the diversified demands of various upper applications for computing resources and computing resources. This paper introduces the formation mechanism, technical architecture and functional composition of heterogeneous AI computing power operation platform, and suggests that heterogeneous AI computing power operation platform could be optimized and developed from the three aspects of technical ability improvement, application ecology improvement and standardized evaluation.
In this paper, the “Five Forces of Computational Power” is proposed to comprehensively evaluate the computational power of data centers. The model solves the problem that the measurement units of different types of indicators are not unified in the process of computational power evaluation, and combines the bidirectional projection method and TOPSIS method to evaluate the comprehensive situation of computational power more scientifically and effectively. Finally, an example is given to illustrate the validity and practicability of the model.
In the era of interconnection of all things, the demand for computing power and punctuality of massive terminal devices is increasing, which effectively promotes the development of edge computing. However, the disadvantages of limited computing power resources and narrow computing power coverage scene of edge computing make it difficult to analyze and process the global data and form a long-term storage of data. Cloud side collaboration gives full play to the advantages of cloud computing and edge computing, realizes the collaboration of cloud side nodes in resources, data, business and other fields, and can better meet the computing power needs of the Internet of things era. Through the analysis of the cloud edge collaborative construction scheme of large central enterprises, this paper summarizes the technical highlights and reference significance of cloud edge collaborative construction of large central enterprises, which will provide important support for the practice of cloud edge collaborative technology of Chinese enterprises.
With the continuous expansion of computing power demand scenario, the form of computing power facilities and computing power supply has changed. The trend of network cloud transformation is obvious. The deep integration of cloud networks can quickly realize the interconnection between clouds and users, and the efficiency of cloud services is improved. In recent years, ubiquitous terminal IOT scenarios become increasingly rich, users need ubiquitous access, on-demand, customized, efficient and convenient computing services. The computing network collaboration system develops rapidly, and the popularity of computing network is increases. The computing power network improves the quality and efficiency of computing power service through technical means such as network cloud and pool transformation, unified identification of computing network resources, intelligent arrangement of computing network resources, data minimalist forwarding and so on. The construction of computing network collaboration system needs the coordination and linkage of data center and network facilities. This paper discusses the impact of computing network collaboration on computing power industry, and gives development suggestions.
Based on data reusability and low-carbon computing, this paper redefines the term data for algorithm. With techniques like data preprocessing, pre-training, desensitization and encryption, a proper standard system is built to produce normalized ways of management and application. According to the properties and values of data for algorithm, this paper introduces FAIRDNS principle and offers suggestions on its research boundary, classification and storage methods. Then, it provides the thoughts about standardization, repository and ecosystem.
As the power density of data centers increases, heat dissipation becomes a major challenge to ensure the smooth running of devices and promote green and healthy development of data centers. Liquid cooling technology can effectively solve the problems of high-density deployment and partial overheating of the machine room, among which immersion liquid cooling technology has been widely considered because of its outstanding heat dissipation and energy saving advantages. In this paper, two typical immersion liquid cooling modes of single phase and phase change are introduced, the key issues of the coolant under the two modes are discussed, also the performance and application differences of hydrocarbon, organosilicon and fluorocarbon compounds, as well as their impacts on safety and environment are emphatically analyzed.
The datacenter is the database to the digital era, has shown a trend of rapid development in recent years, and the energy consumption problem of the datacenter industry is becoming more and more serious. In order to improve the overall utilization efficiency of the datacenter, this paper is based on the principle of optimal granularity design, takes the optimal particle size design scheme of a decentralized refrigeration system as an example, analyzes the precise matching of heat dissipation and refrigeration, thereby expands the fine adjustment of the data center facility level, and puts forward the best granularity thinking on architecture design, working condition selection and energy selection about the datacenter computing efficiency, water efficiency and carbon efficiency.
UPS battery is an important component to ensure the stable operation of the data center, and its safety is related to the reliability and stability of the data center power system. In recent years, the monitoring, operation, and maintenance management of batteries have attracted the attention of academia and industry to avoid the occurrence of power accidents caused by battery failure. Three development stages of operation and management of UPS battery in the current data center are analyzed. An intelligent operation and management system is proposed, integrating the monitoring system, flexible alarm, asset management, and fault prediction. The algorithms for batteries’ health monitoring, anomaly detection, and performance prediction are further summarized.
Supercomputing is the embodiment of a country’s comprehensive national strength, one of the key technologies supporting the country’s sustainable development, and occupies an important strategic position in national economic construction, scientific research and national defense security. In recent years, China’s supercomputing industry has ushered in a stage of rapid development, and supercomputing has gradually become one of the important means to promote social and economic sustainable development, industrial transformation and upgrading and improve people’s living standards. This paper introduces the current development status of supercomputing industry at home and abroad, points out the development value and main problems of supercomputing industry in China, and puts forward development suggestions from different aspects.
Under the national digital economy strategy, data has become an important national asset as a production factor for developing digital economy. Data storage, as a carrier for carrying data, is also the focus of the 14th Five-Year Plan. This paper proposes to promote data storage transformation by considering the ‘Dual carbon’ goals and big data strategy, and actively explore software and hardware technologies and solutions that minimize energy consumption while ensuring data storage performance. This will provide strong support for the sustainable and stable development of the digital economy.
Green new materials are the important basis for all industries to achieve carbon peak and neutrality. This article summarizes the key directions and development layout of low-carbon and environmental materials in foreign countries’ new material development strategic plans and the relevant policies of green new materials in China’s new material planning since the 12th Five-Year Plan. The industry chain structure and development status of light alloy, new energy materials, green building materials and advanced chemical materials, which play an important role in the process of realizing the goal of carbon peak and neutrality in China, were studied and summarized. The suggestions were put forward according to the development difficulties and pain points of each key direction.
Digital transformation has become the key to improve the coverage, precision, and sustainability of inclusive financial services for small and medium-sized banks, and will build new advantages for the development of inclusive financial services for small and medium-sized banks under the new development pattern. However, the development of digital inclusive finance for small and medium-sized banks still faces many challenges in data, technology, information, risk control, and talent. Through the digital transformation and development in areas such as data governance system, the underlying technology of science and technology, financial information security, risk management ability, and digital technical talents, we will promote the digital inclusive finance of small and medium-sized banks to empower the real economy and better realize the inclusive regional economy and high-quality growth.
Based on the analysis of the characteristics and limitations of trusted relay Quantum Key Distribution(QKD) network, a quantum key service platform technology based on QKD network virtualization is proposed. Quantum key service platform, which is an innovative scheme to solve the practical problemsfaced by QKD, has higher security, service efficiency and application flexibility, and has good application prospects.
How to manage the huge amount of unstructured data and extract greater value from it determines the direction of the second half of the global economy. This paper discusses the causes of unstructured data management, management methods and the development of unstructured data management in the context of the current situation of global data production and storage as well as the development of data management theories and technologies, and provides an outlook on the construction of intelligent digital management in the future.